The sensitivity of Austrian forests to scenarios of climatic change:: a large-scale risk assessment based on a modified gap model and forest inventory data

In a spatially explicit climate change impact assessment the modified patch model PICUS v1.2 was applied to simulate the transient response of current forests in Austria under three climate change scenarios which were based on regionalized GCM-scenario data. The forest model was initialized with ground-true stand and soil data from more than 2800 sample plots of the Austrian Forest Inventory (AFI). A comparison of simulated equilibrium species composition under current climate and expert reconstructions of PNV at the sample plots of AFI showed that the model responded realistically to the spatial variability of soil and climate characteristics. In deriving potential climate change impacts the simulation under current climate was used as a reference. Impact criteria representing the period 2000-2050 and long-term criteria derived from simulated site-specific potential natural vegetation (PNV) were used in a multiple-criteria approach to calculate short-/mid-term as well as long-term climate change impact indices. The study showed that neither transient short-/mid-term nor long-term PNV-based indices alone are sufficient to indicate the possible consequences of climate change on existing forests. Based on the results of the study the combined use of such climate change impact indices is recommended. A major finding was that beyond a temperature increase of approximately +1degreesC (no changes in precipitation) the proportion of inventory plots showing severe climate change impacts increased markedly. While at higher elevations under warmer climates the set of suitable tree species increased due to increased competitivity of broadleaved species, the study suggests that under the set of analysed climate change scenarios at low-elevation sites Picea abies would become unsuitable as a crop species. Limitations of the presented approach are discussed and conclusions regarding possible consequences for forest management are drawn. (C) 2002 Elsevier Science B.V. All rights reserved.